Links for 2022-10-27
Why Weren't Hot Air Balloons Invented Sooner? https://www.lesswrong.com/posts/bvqC4Ci7rXq4sN9df/why-weren-t-hot-air-balloons-invented-sooner
Weird metal that's also glass is insanely bouncy https://www.youtube.com/watch?v=QpuCtzdvix4
"Results from a sample of 1,365 individuals suggest that incentives in the range of 8-125 USD backfire, inducing lower vaccination willingness than simply offering vaccines for free." https://www.nottingham.ac.uk/cedex/documents/papers/cedex-discussion-paper-2022-14.pdf
How much do CEOs matter? Using the entirety of Danish firms from 1994 to 2003, the authors use the sudden death of a CEO, or a close family member, to estimate the marginal value of CEOs. At a minimum, 11% of profitability is due to CEOs alone. https://www0.gsb.columbia.edu/mygsb/faculty/research/pubfiles/3177/valueceos.pdf
The US Navy wants swarms of thousands of small drones [MIT Technology Review] https://archive.ph/A7Nkl
How AI will change everything on the internet https://marginalrevolution.com/marginalrevolution/2022/10/how-ai-will-change-everything-on-the-internet.html [people starting to notice that AI is going to be transformative]
“Why are snowflakes symmetrical? How can ice crystallizing on one arm 'know' the shape of the other arms on the flake? Nature adds one water molecule at a time. The molecules always try to choose the most energetically favored position on the frozen body. Because these laws of creation of a snowflake are symmetric with respect to the rotational symmetries, it follows that any symmetry that exists at the beginning - a hexagonal symmetry of a small number of molecules in the initial crystal - will be preserved. It's pretty much inevitable that all the arms are growing approximately equally, so the initial symmetry group is preserved and becomes a symmetry of a macroscopic object.” https://physics.stackexchange.com/questions/3795/why-are-snowflakes-symmetrical
“Transformers have shown remarkable capabilities - but can they improve themselves autonomously from trial and error? New research shows how a single transformer can explore and solve tasks using this method without ever updating its parameters.” https://arxiv.org/abs/2210.14215
Finetuning GPT-2-0.7b to better stories than GPT-NeoX-20b https://arxiv.org/abs/2210.07792
“This simple strategy enables the open-source GPT-J-6B model to match and exceed the performance of few-shot GPT3-175B on 15 of 20 popular benchmarks” https://arxiv.org/abs/2210.02441
Self-consistency amazingly improves chain-of-thought prompting, even 50+% improvements on many tasks. https://arxiv.org/abs/2203.11171
I wrote something related to this a few years ago:
A rarely mentioned side effect of superhuman artificial general intelligence is that, even if it doesn't kill us, it will remove almost all meaning. Everything anyone cares about will be literally one prayer away from being fulfilled.
For example, what if you came up with a philosophical conundrum? Well, just ask God to solve it for you. And if you're not smart enough to understand the solution, just ask God to make you smart enough.
Or what if you wanted to do mathematics? You could trivially integrate the resources of a specialized Matrioshka brain into your consciousness and implement and run an ideal mathematician.
Everything you could do has either already been done or can be done better by God.
But surely, you wonder, there must be fantastic virtual environments to explore. And what about sex? Well, God thoroughly understands what it is that makes exploration and sex fun for humans. It knows how to implement the ideal adventure in which you save people of maximal sexual attractiveness and could instantly integrate the memory of such an adventure for you, or simulate it a billion times in a few nanoseconds. And the same is true for all possible permutations that are less desirable.
But the consequences are even deeper than that. Concepts such as creativity or fun will be perfectly understood as mechanical procedures that God can easily implement and maximize. For God, human happiness is conceptually no more interesting than an involuntary muscle contraction. For God, the incomprehensible complexity of your human values is a conceptually simplistic, and transparent set of rules, barely more interesting than a rat pressing a lever in order to receive a short electric stimulation of its reward center.
In summary, artificial general intelligence is literally the last discovery we have to make. At that point, the universe has understood itself.
The movie has been watched.
The game has been won.
The end.